Learn 50 Data Analyst Concepts In 6 Minutes
2 Video Views·Dec 19, 2024
Industrial Technology
I'm going to give you 50 data analyst concepts in just 6 minutes.
Keep in mind that you don't need to remember every single thing, just having heard about
them will be very useful.
Now in the beginning you will recognize some of these, but I guarantee that you won't recognize
them all.
Let's get started.
Data collection means gathering data from various sources.
It's like hunting for valuable information that you can use.
Data cleaning is all about tidying up the data, fixing mistakes, removing duplicates
and getting rid of everything we don't need.
Data science is the general term used for the study of data to extract insights for
example of a business.
Data analysis is often seen as a part of the data science field.
Data transformation means changing data into a more suitable format that we can work with.
Data modeling is like creating a map out of our data to see how our data looks like and
how parts are connected with each other.
Data integration means putting all of our data together in one place.
Data visualization is all about our way of showcasing data in a visual way to help people
understand what's going on and also to showcase our findings.
Inferential statistics is a way of using a smaller sample to try to get an idea of the
whole picture, like trying to understand a forest by just looking at a specific section
of the forest.
This is very useful because it's super hard to analyze a whole forest.
Probability is a concept that we all know, it's how likely it is that something is going
to happen.
A hypothesis is the prediction we make about something.
If we test our hypothesis to see if we're correct, it's called hypothesis testing.
Correlation is a way to measure how two things move in relation to each other.
Make sure that you don't confuse it with causation, which is if it directly causes it.
Regression analysis is a method used to find the pattern and understand the relationship
between a dependent variable and one or more independent variables.
For example, we might use a person's height to predict their shoe size, and this is a
really important concept in data science.
The next concept is outlier.
An outlier is a data point that is significantly different from other observations.
It could be an error or real, but a unique occurrence.
Time series analysis is where we look at the data collected over time and try to spot trends
over time, just like in the name time series analysis.
Machine learning is where we train computers to learn from data and make decisions or predictions
like teaching a robot how to think.
Predictive modeling is all about creating models that predict future outcomes based
on past data.
It's basically using past experience to predict the future.
Clustering is where we group similar items or data points together, like all customers
that shop only on Tuesdays.
Classification is where we categorize new data based on past data, like sorting new
data into separate buckets that we already created.
Data mining is finding patterns or insights in large datasets.
Text mining is basically the same, but you extract useful information from a larger text.
It's like reading a book and writing down the most important parts.
Data warehousing.
We store and manage a large amount of data in a data warehouse because it is efficient
and easy to use.
It also allows for more complex data to be stored than in just a database.
Now, this video is sponsored by no one, so let's just keep going.
Big data refers to huge datasets that are so big that they require specific big data
tools to be able to work with.
Structured data means data that's organized in a specific way, and often that means that
it's stored in a spreadsheet with rows and columns.
Unstructured data has no clear structure to it, and it's usually much harder to store.
You can often not just put it in a spreadsheet, so it's harder to both store and to work with.
Semi-structured data is in between both of them.
It could, for example, be stored at a first level, but it's still slightly unstructured
if you do take a look.
Data quality is all about how accurate and reliable our data really is.
Bad data quality is essentially bad data, which means bad decisions.
Data governance is a term that refers to how we manage and protect our data, for example
in the company.
Data privacy is pretty self-explanatory.
It's basically keeping personal information in our data secure, which is really, really
important.
Both for privacy reasons, but also due to specific laws.
Data security is keeping the data itself safe, for example from hackers and other people
that are trying to access the data and steal it.
ETL stands for extract, transform, and load, and is a way to move data from multiple sources
and store it into one place where we can use it.
It's a must-know term for data engineers, but it's still really good to know as a data
analyst.
Data marts are smaller subsets of a data warehouse, and for example in a company, a specific department
might only need to see and access a section of the data that's relevant to them.
Next, let's talk about data lakes.
They're a big storage system with raw data in the original format, in a pretty chaotic
way, like a big lake of unprocessed data.
Data is moved through a data pipeline, and it's processed along the way so that we can
move it.
For example, data from Twitter into your database.
A data schema is a blueprint showing how data is organized and structured, like a floor
plan for a house, but in this case for data.
Normalization is a way of organizing data into a database effectively.
On the other hand, denormalization is a way of optimizing your database to improve performance
by adding extra, redundant data.
It takes more storage and comes with both pros and cons.
A primary key is a really important term for example in SQL and refers to a unique identifier
for each record in a database.
It's basically like a unique identifier for each row in a spreadsheet, like customer ID.
A foreign key is a primary key in another table, but not in the one that you're currently
referring to.
It can be used to connect different tables, for example.
Indexing is pretty self-explanatory.
You can use an index to speed up data operations in your database because you don't have to
search for them in the same hard way, because it makes it way faster to, for example, search
up the data.
SQL stands for Structured Query Language and is a programming language used to work
with databases.
You describe what you want to do and SQL does it for you, so you don't need to know what.
NoSQL is a type of database that can handle a lot of different types of data, not just
neatly organized tables, making it great to work with less structured data.
Quite funny, data analytics, you're looking through data to find patterns or insights,
which can help make decisions or solve problems.
Business intelligence, or more commonly referred to as BI, is about taking data and turning
it into useful information to help businesses do better.
A dashboard is like the control panel in your car, but for data.
It shows you the most important information all in one place so that you can easily see
how things are going, usually in real time too.
KPI stands for Key Performance Indicator and it's a type of measurement that shows how
well something is doing.
For example, a business might look at sales numbers as a KPI to see how well they're performing
at the moment.
Real-time analysis is when you take a look at the data as it comes in and you directly
analyze it right away.
Data wrangling is like cleaning and organizing your data, so it's easier to work with, basically.
Pandas is a very common Python library used for data manipulation and data analysis that
I think everybody should be aware of if they're a data analyst.
Matplotlib is a plotting library for Python also used to create data visualizations.
And finally, a histogram is a visual representation of a distribution of a dataset.
Now this was only the beginning.
Check out this video next if you really want to become a data analyst and I'll show you
exactly how to do that.

